Paul Schlie wrote:
> then presumably scheme is broken defining:
>
> (fl= +nan.0 +nan.0) => #f ; being that ieee defines it as #t
As John Cowan pointed out, the IEEE-754 standard
requires equality tests that involve a NaN to
come out false.
> or
> (flexpt -1.0 +inf.0) => NaN (hopefully); as ieee defines it as oddly 1.0
The IEEE-754 standard does not define anything
corresponding to Scheme's expt or proposed flexpt.
A full (but unofficial) text of the now withdrawn
IEEE-754 standard is available online, as is the
full text of the current working draft that has
been proposed as its replacement, IEEE-754R [1].
Will
[1]
http://www.validlab.com/754R/
Received on Sun Oct 08 2006 - 20:35:56 UTC